Your brand has no faces, and giving it some feels like losing control.
Everything the company publishes is correct and lifeless. The content that actually travels in your market is written by individuals with an opinion, and none of it comes from your company. Letting twelve people speak freely risks twelve tones of voice. That trade-off is the reason most B2B brands stay on the company page.
€ 299/month flat · 5 seats included · Cancel anytime · 90-day ROI guarantee
Brand Wizard
Define your company's voice & DNA.
Part 1: Tone & Guardrails
Brand Adjectives
Blacklisted Words
Part 2: Company DNA
Elevator Pitch
We help B2B marketing teams turn their employees into authentic brand ambassadors on LinkedIn, combining AI content generation with strict brand governance to ensure every post is both personal and compliant.
How do you let employees post on LinkedIn without losing brand consistency?
Consistency is decided before anyone writes a word, in what every draft is built on. Reviewing finished posts is too late to produce it and too slow to sustain. Four things have to be true at the same time:
- The substance is shared, the voice is not. A company owns its position, its vocabulary and the problem it claims to solve. The person writing owns tone, examples, point of view and which stories get told.
- Guardrails have to survive the second draft. Rules that apply when a draft is generated and then fall away the moment someone asks for a rewrite are decoration.
- The check happens during writing. Anything that catches a problem after publication is a correction, and corrections are what people remember about the program.
- The rules are visible to the people they apply to. A blacklist nobody can see is a trap. The same terms that steer the drafting are flagged in the editor, so the author knows the boundary while they are inside it.
Disclosure before we go further: we build Authira, one of the tools in this category. The four conditions above hold whether you buy anything or not.
None of them requires an approval step, and that is the design decision worth stating early. Authira has no review queue, and an admin cannot read a draft before it is published. A post that waits for sign-off arrives late and teaches its author that publishing is admin work. If mandatory sign-off is the rule in your organisation, you can stop reading here and buy an enterprise suite instead. The last section of this page lists everything else we deliberately do not do.
Four ways out of the trade-off, and what each one costs
The choice is usually framed as human against controlled, which makes it unresolvable. It is more useful as four working arrangements, two of which you are probably running already.
| Approach | What it buys | What it costs |
|---|---|---|
| Publish from the company page only | One voice with nothing to coordinate, fully under your control. | Reach that follows a page rather than a person. Followers accumulate and the posts sit still. |
| Everyone reshares the same company post | Consistency, and one message carried further than the page carries it alone. | The same text appearing under several names, which readers recognise quickly and stop reading. |
| Everyone posts freely, no shared rules | Real voices, and the reach that comes with them. | No shared position. Drift is only visible after publication, and the only correction available is a conversation. |
| Shared substance, separate voices | Twelve people who sound like themselves and argue the same thing. | Somebody has to write the brand rules down once, properly, before anyone drafts anything. |
The first row is where the budget usually sits, and it is the one with the clearest evidence against it. Followers on a company page accumulate while reach does not follow them, which is why boosting a post buys impressions without buying attention. There is a second cost to that row that rarely makes it into a deck. Nobody joined marketing to produce wallpaper, and the distance between the work you wanted to make and the work that survives review is the part that wears people down.
The second row is the arrangement a lot of advocacy tooling was built around, and it holds up until readers notice that four people published the same paragraph on the same morning.
The fourth row is the one worth the setup cost, and the cost is real. Writing down what your brand actually argues, in a form specific enough to steer somebody else's draft, takes an afternoon and a decision or two you may have been avoiding. It is also the only one of the four where the result arrives as a number rather than as an impression, which is what makes it defensible at budget time.
Same substance, different voices
A thought leadership program that runs on twelve people needs two layers behind every draft. One belongs to the company and is identical for everybody. The other belongs to the individual and is different for everybody. They are assembled together each time somebody generates a post, and the section after this one shows what that produces.
What the company holds
Brand adjectives that set the register, the elevator pitch, the core problem the business solves, and a blacklist of terms nobody may use. If you already have a tone of voice document, all four are usually in it somewhere, so this is extraction rather than invention. You edit them as the admin and they apply to everybody from the next draft on.
What the person holds
A tone setting on a scale from factual to approachable, which they choose themselves. Expertise topics and stored anecdotes come from four questions at setup. Red lines and the German Du or Sie choice they add by hand.
The part worth checking in any tool you evaluate is what happens on the second pass. Both layers are attached again when a member asks for a rewrite of a paragraph or for alternative opening lines, which is the point where guardrails usually leak, because a rewrite is a new request to the model and the company rules have to be sent with it.
Term matching is built for German as well as English, so a blacklisted stem is still caught inside a compound and across inflected forms. How the same rules function as policy enforcement instead of brand coherence is covered on the compliance guardrails page. What the drafting itself looks like for a member sits on the content assistant page.
The same argument in two voices
What the two layers produce is easier to see than to describe. Below is one company position, held by two people with different tone settings and different stored anecdotes.
Tone set toward factual and precise
A discount request that arrives before the scope is agreed is usually a question about sequence. While the deliverable is still open, any price is an estimate, and discounting an estimate produces a smaller estimate. We now hold the price conversation until the scope document is signed. Clients tend to find that easier, because by then they are agreeing to something specific.
Tone set toward casual and approachable
"Can you do something on the price?" That question almost always turns up before we have agreed what we are actually building. Which is fair enough, nobody wants to sign a blank cheque. The catch is that while the scope is still fuzzy, my number is a guess too. So we swapped the order. Scope first, signed. Price after. People are noticeably more relaxed about a figure once they know exactly what it buys.
Both posts make the same claim and argue the same position on how the company sells. What differs is the register and the story each one reaches for, because the two profiles carry different tone settings and different anecdotes. Both texts were written for this page to show the shape. Neither is a customer post and neither is a benchmark.
The objection worth taking seriously: will it sound like AI?
You have read enough machine-written LinkedIn posts to recognise the cadence, and your brand is what gets damaged if your people start publishing it. The honest answer here is a mechanism, not an adjective, and it runs in three stages.
Ten rules name the patterns that give machine text away. They are part of the instructions behind every generation, and they apply in all five languages the product writes in. They cover punctuation habits, specific sentence figures, a set of transitions, and the closing moves that read as generated to anyone who spends time on the platform.
Eight of those are checked again by code. After a draft comes back, a deterministic pass reads it paragraph by paragraph. No model is involved at that stage, so the result is reproducible and every rule in it is covered by tests. If nothing is flagged, nothing further happens and the draft goes straight to the author.
Flagged drafts get a repair pass with hard limits. The repair stage sees the whole draft, and returns only the paragraphs it changed. Those are swapped in by exact string replacement, so every paragraph it did not name stays byte-identical to what was generated. Its instructions forbid inventing anything that was not already there, and numbers and proper nouns have to survive unchanged.
Two honest limits. The detector is deliberately tuned to over-flag, which means the model supplies the judgement and a mechanism is not a guarantee. And every draft stays editable by the person publishing it, which is the last and best filter. The same rules govern the copy on this website, which is the only test of them we can show you in public.
A number you can take into a budget review
Brand work is the first budget questioned and the hardest to defend, because awareness is not a figure anybody respects. Reach converted into euros is at least a figure of the same kind as the one you are defending it against.
Team impressions are converted at a fixed € 34 per thousand and reported as earned media value, alongside posts published and the share of members who were active. The period runs to any of the last twelve months and exports as a PDF. The reporting screen is where that lives. It is a flat rate with no trust multiplier and no engagement weighting, and the sourcing behind it sits on the earned media value calculator, which uses the same figure. Running your team's current reach through it before you buy anything is the cheapest version of this argument.
Be precise about what the number is when you present it. It answers what the equivalent paid reach would have cost. It says nothing about pipeline or revenue, because there is no click tracking and no attribution in the product. Presented as reach it holds up. Presented as revenue it will be taken apart by the first person in the room who asks how it was calculated.
Management Report Preview
June 2026
Network Growth: Last 6 Months
What you are actually asking of your people
This is the part that is easy to underestimate from the brand side. Resharing a company post costs a colleague nothing. Putting their own name and their own network behind a position the company holds is a much larger request, and a much more valuable one, and it changes who belongs in the program. The people worth inviting are the ones with something to say about the work, which is rarely the same list as the org chart.
What it costs them week to week is small and worth stating plainly. Drafting and editing happen in the assistant, and the finished post reaches LinkedIn by copy and paste from their own account, because nothing here publishes on anybody's behalf. The recurring task is the analytics upload. Each person downloads their own LinkedIn export and uploads it, since there is no API connection. A member who drafts and never uploads is invisible in your reporting, which is the usual reason a dashboard looks emptier than the program is.
One thing to decide before the invitations go out. The team view is visible to every member rather than to admins only. It carries collective reach, the monthly target, consistency streaks and a leaderboard with colleagues' names on it. For most teams that visibility is the point of the exercise. For a few it is a conversation worth having first.
What Authira does not do
Worth knowing before you evaluate rather than after. These are design decisions, not gaps waiting to be filled, and any one of them is a good reason to buy something else.
- No approval workflow, and no admin view of drafts. This is the sharpest limit in the product for a brand-protective buyer, so it goes first. Nobody reviews a post before it is published, and unpublished drafts are readable only by their author, enforced in the database rather than by a setting. Guardrails are preventive instead: your blacklisted terms are flagged while somebody writes and again before they mark a post ready, and the author decides. If sign-off before publication is mandatory in your organisation, this is the wrong tool.
- No scheduling and no auto-posting. Finished drafts are copied into LinkedIn by the person who wrote them.
- No LinkedIn API connection. Analytics come from each member uploading their own export and keeping it current.
- No click tracking or pipeline attribution. The product reports reach and earned media value, never revenue or website traffic.
- LinkedIn only. No other network, and nothing for paid social.
- No company content library to reshare. Drafts start from the individual. If a curated feed your team can amplify is what you came for, that is a different product category.
- It is not an agency. Nobody at Authira writes your posts or runs the program for you.
- Built for roughly 3 to 30 participants, self-managed. Past about 30, a dedicated program manager and enterprise administration start to matter, and we compare that market openly.
€ 299/month flat · 5 seats included
5 seats included · Cancel anytime · 90-day ROI guarantee
See full pricingBrand consistency and employee posts: FAQ
How do twelve people post in their own voice without ending up with twelve different positionings?
By separating what the company owns from what the person owns. On the company side sit brand adjectives, an elevator pitch, the core problem the business solves and a list of words nobody may use. On the person's side sit a tone setting, expertise topics, stories they have lived and terms they personally will not use. Both are attached to every draft the assistant produces, and both are attached again when someone asks for a rewrite or for a different opening line. What varies between two members is register and example. What does not vary is the position being argued.
Will everyone end up posting about the same thing?
No, because the shared layer is a position rather than a topic. Two people writing from the same brand rules can write about a client conversation and about a technical decision in the same week, and both posts still argue what the company argues. Topics come from each person's own expertise tags and their own stored anecdotes, so a specialist and a founder do not draw from the same well. If two members do choose the same subject, they get different posts, because the tone setting and the anecdote behind each one differ.
Can we use our existing tone of voice guidelines?
Partly, and the translation is short. Authira stores brand adjectives, an elevator pitch, the core pain point your business solves, and a blacklist of terms. A tone of voice document usually contains all four in prose, so setting it up is a matter of extracting them rather than inventing them. What does not carry across is anything that needs a human interpretation, such as a rule about when irony is appropriate. Guidelines of that kind still belong in the conversation you have with the people writing.
What happens if someone posts something off-message anyway?
The same thing that happens today, which is a conversation. Authira works preventively instead of by review. Blacklisted terms are flagged while the person is writing, and once more when they move a post to ready, and the author decides. There is no queue that holds a draft back and no admin who can veto it. That is a design decision rather than a missing feature, and if your organisation requires sign-off before publication, an enterprise suite with a review workflow is the correct buy.
Our people are not writers. Is this realistic for them?
Most of them will not be, and the product assumes that. Setup is four questions about their work and one tone setting. After that each generation returns three drafts to choose between, with an editor that offers alternative opening lines and shorter or sharper rewrites of a selected passage. The work left to the person is choosing which draft sounds like them and correcting the parts that do not. Willingness to be visible under their own name is the real constraint here, and it is worth testing before you buy seats for people who have not agreed to it.
Can I see what my team is drafting?
No. Drafts are visible only to the person who wrote them, and that is enforced in the database rather than by a setting an admin could change. As the admin you see the brand rules, the team reporting and who has been active. You do not see unpublished text. For a brand-protective marketing lead this is the sharpest limit in the product, and it is better discovered here than after a purchase.
Brand budget is the first thing questioned. Does this give me something to defend it with?
It gives you reach expressed in euros, which is a different argument from awareness. Team impressions are converted at a fixed € 34 per thousand and exported as a PDF. Be precise about what that number is when you present it. It answers what the equivalent paid reach would have cost, and it says nothing about pipeline, revenue or attribution, because Authira does not track clicks or conversions. Presented as reach it is defensible. Presented as revenue it will be taken apart in the meeting.
Is this just an AI writing tool?
Drafting is the part a single person could get elsewhere, and it is table stakes in this category. What a creator tool cannot do is hold a company layer across several people: one set of brand rules and one blacklist applied to everyone's drafts, a shared monthly target, a team view everybody can see, and reporting that aggregates the whole group in euros. If one person needs to write better LinkedIn posts, a creator tool is cheaper and sufficient. The team layer is what this is for.
Keep reading
- How brand guardrails are enforced: the same rules seen as policy enforcement, including where they stop.
- The LinkedIn EMV calculator: put a euro figure on the reach your team already produces, before you buy anything.
- Why employee advocacy programs fail: the four failure modes, and why collapse is a design problem rather than a people problem.
- Running a program without software: if you already have people posting, what a tool actually takes off your desk.
Let the company sound like the people in it
You write down what the brand argues once. Everybody drafts inside it, in their own voice, and the result comes back as reach in euros. Flat € 299 per month, 5 seats included, cancel any time.
5 seats included · Cancel anytime · 90-day ROI guarantee
Last updated